Abstract:With the increasing demand for process safety and production efficiency, many research efforts have been made for nonstationary process monitoring in recent years. However, existing methods usually neglect dynamic characteristics of industrial processes, which may lead to misleading results. Only few literature studies address the problem of nonstationary dynamic process monitoring, and most of them are mainly based on the assumption that nonstationary variables are integrated with the same order I = 1. In order to deal with the general case, where nonstationary variables are integrated with different or higher orders, dynamic stationary subspace analysis (DSSA) is proposed in this paper. In DSSA, the time shift technique is introduced to model dynamic relationships, and an optimization problem is described to estimate the stationary projection matrix similar to stationary subspace analysis (SSA). Different from traditional SSA, the alternating direction method of multipliers is utilized to solve the optimization problem, and detailed iteration expressions are derived. After the stationary projection matrix is obtained, the Mahalanobis distance is adopted for monitoring stationary components of augmented data. The monitoring performance of DSSA is demonstrated by case studies on a simulated nonstationary dynamic process, a nonstationary continuous stirred tank reactor, and a practical ultra-supercritical power plant.

Study on the two-stage least square algorithm and its performance for nonstationary stochastic system

Novel parameter identification algorithm for nonstationary stochastic system

Combined Algorithm for Time-Varying Systems Based on the Robust Min-Max and the Damped Least Squares Estimation

Exact Stationary Solutions Independent of Energy for Strongly Nonlinear Stochastic Systems of Multiple Degrees of Freedom

Online Successive Convex Approximation for Two-Stage Stochastic Nonconvex Optimization

Using Svm To Model And Control Nonlinear Dynamical Systems

Nonnegative Least-Mean-Square Algorithm

Stochastic Behavior of the Nonnegative Least Mean Fourth Algorithm for Stationary Gaussian Inputs and Slow Learning

Steady-State Performance of Non-Negative Least-Mean-Square Algorithm and Its Variants

Optimal Step-Size Update Equation in Nonstationary Environment and OVS-LMSII Algorithm

Convergence of two-stage multisplitting method with ssor multisplitting for linear systems with H-matrices

EXPLICIT ITERATION ALGORITHM FOR NON-STATIONARY DYNAMICS RESPONSE ANALYSIS OF EXPLICIT ITERATION ALGORITHM OF NON-STATIONARY STOCHASTIC DYNAMIC RESPONSE ANALYSIS BASED ON EQUIVALENT STOCHASTIC LINEARIZATION

Study of on-line weighted least squares support vector machines

Nonstationary Process Monitoring Using Sparse Stationary Subspace Analysis.

Learning Stable Linear Dynamical Systems with the Weighted Least Square Method

On adaptive stochastic extended iterative methods for solving least squares

Distributed Least Squares Algorithm for Continuous-time Stochastic Systems Under Cooperative Excitation Condition

Dynamic Stationary Subspace Analysis for Monitoring Nonstationary Dynamic Processes

A Weighted Least-Squares Method for Non-Asymptotic Identification of Markov Parameters from Multiple Trajectories

Stochastic Regularized Newton Methods for Nonlinear Equations

Two Iterative Algorithms for Stochastic Algebraic Riccati Matrix Equations.